Durum denklemleri regresyonu kullanilarak retrograt yoǧuşma bölgesinde hacimsel tahminler

A study on the probable bias on the volumetric predictions of a regression-tuned EOS was carried out using different sets of volumetric data for a well-defined complex gas condensate system. Components of the recombined gas condensate composition included nitrogen, CH₄, ethane, propane, n-octane, n-decane, n-dodecane, etc. The Peng-Robinson (PR) EOS predictions might be satisfactory for many simpler systems. For well-defined condensate fluids, volumetric predictions could be inaccurate and show process specific trends. The volumetric shifts method could not correct the overall shape of the liquid drop out curve alone. Adjustment of the critical properties of the heavier ends should be considered along with the CH₄ and heavier fraction binaries for well-defined systems. The PR EOS could not properly account for the intermolecular forces during a CCVD process. Tuning with a specific volumetric data set imposed bias on the predictions of the other volumetric processes. Simultaneous tuning with the multiple data sets did not necessarily enhance the predictive ability of the EOS. The biased predictions were not artifacts of the tuning techniques utilized. The regression based tuning could be a fast and accurate procedure but must be applied cautiously.